Weekly Distribution of Monthly Sales Forecasting

In most cases, the Sales Forecasting is performed on a monthly basis for many reasons. The basic reason is that the monthly forecasts tend to result in higher accuracy and this is what every forecaster seeks to. In addition, forecasting for shorter time buckets tends to be much more difficult.When I started doing some forecasting myself I soon realized that I had to shorten the time bucket. The initial monthly forecast should be distributed over weeks and then the accuracy should be measured. If the accuracy was about the same as on the monthly forecast, which is very unlike to happen, then it would be safe to say that the demand is constant over the month.

Although it seems easy at first, it is a really tricky procedure. Every month has different number of weeks and each week has different number of working days. It’s a really bad idea, just to divide every month with 4 and try to distribute the forecast this way.
Excel has a lot of date functions that can help you perform this, such as the Days360, NetWorkDays, WeekNum, etc, but still this task is not very easy. Apart from this, you should know that Excel does not calculate the ISO Week number as most calendars do (http://en.wikipedia.org/wiki/ISO_week_date)
but the arithmetic week. You have to make your own custom ISOWeek function to calculate it.
Taking under consideration the above, I decided to write a useful Excel Add In, to make my life easier.